Surrogate parameter identification for coupled lithium-ion cell models

Abstract

Listen

Translate article

Purpose – The need to optimize and simulate lithium-ion cells is of high importance to their durability. It is possible to use semi-empirical models based on equivalent circuits to model the dynamic behavior. Still, parameter identification is a time consuming task if many equivalent circuits are coupled into a finite element-based simulation. The paper aims to discuss these issues. Design/methodology/approach – In this paper the authors present an approach to estimate parameters in a very efficient way by using a single equivalent circuit model as a surrogate model on measurement data. The results of the surrogate model are periodically linked back to the original complex model via the so called space mapping method. The authors validate the approach and compare it to the original problem. Findings – As a remarkable result, the authors report the achieved reduction of computational cost by approximately 87 percent, which equals a speed up factor of 8. Originality/value – To the best of the authors’ knowl...